We examined wetland selection by breeding Green-winged Teal (Anas crec
ca carolinensis) in terms of wetland fertility near Riske Creek, Briti
sh Columbia, Canada. We conducted 12 breeding-pair surveys to classify
use of 96 wetlands. We evaluated wetland fertility by analyzing water
chemistry (pH, conductivity, alkalinity [CaCO3], total phosphorus, ch
lorophyll a) and physical characteristics (percent emergent cover, are
a of wetland 0-1 m depth, wetland size). There was significant variati
on in chemical and physical variables between unused (n = 47) and used
(n = 49) wetlands (MANOVA, Wilks' lambda = 0.8054, P = 0.0126). On av
erage, water chemistry measurements, total area (m(2)) and area 0-1 m
(m(2)) depth were greater in used wetlands, whereas percent cover was
lower. Also, there was significant variation in chemical and physical
variables between unused (n = 47), infrequently used (n = 26), and fre
quently used (n = 23) wetlands (MANOVA, Wilks' lambda = 0.6951, P = 0.
0085). On average, water chemistry measurements, total area (m(2)) and
area 0-1 m (m(2)) were positively correlated with wetland use, wherea
s percent cover was negatively correlated. We conclude that wetland fe
rtility is an important factor influencing landscape and geographic di
stribution of breeding Green-winged Teal.